INTRODUCTION

Among electrolyte abnormalities observed in chronic alcoholicpatients, hyponatraemia is relatively common (Kaysen andNorth, 1984; Schaefer et al., 1987; Elisaf et al., 1994).However, the prevalence of hyponatraemia and its underlyingpathophysiological mechanisms in these patients are not welldelineated. In the current study, we attempted to illuminate theresponsible pathogenetic mechanisms of hyponatraemia in agroup of alcoholic patients admitted to our hospital for causesrelated to alcohol misuse.

MATERIALS AND METHODS

We prospectively studied 127 out of 241 consecutivechronic alcoholic patients (120 males and 7 females) aged29–78 years admitted to our hospital for causes related toalcohol misuse over a period of 5 years. To be eligible, thepatients had to have had a large intake of alcohol for at least 5 years, and a weekly alcohol consumption of 600 g or morefor the previous 3 months. The main causes of the patients’admission are shown in Table 1. Patients with diabetes mellitus,renal disease (defined by previous exposure to recognizednephrotoxic drugs, abnormal urinalysis, pathological protein-uria, or creatinine clearance <80 ml/min), ascites, peripheraloedema, acute pancreatitis, chronic obstructive lung disease,recent bleeding from the gastrointestinal tract, septic shock,convulsions occurring 1 h prior to blood sampling, as well as patients consuming drugs affecting acid-base status andelectrolyte parameters such as diuretics, antacids, and potassium,phosphate and magnesium supplements, were excluded fromthe study. In all subjects, psychiatric evaluation was performedon admission and at 1–2 weeks after discharge. There was noevidence of psychotic disorders in any of the patients includedin the study.

On admission, physical examination was performed andvenous blood was obtained for the determination of serumglucose, urea, uric acid, creatinine, total proteins, albumin,

lipid parameters (total cholesterol, triglycerides), osmolality(Posm), potassium, sodium, calcium, phosphorus, chloride,magnesium, and bicarbonate prior to any therapeutic inter-vention. Arterial blood was also obtained for blood gasmeasurements. At the same time, a fresh urine specimen wastested for osmolality (Uosm), uric acid, urea, creatinine, potassium,sodium, calcium, phosphorus, chloride, and magnesium.Standard formulae were used for the determination of thefractional excretion (FE) of electrolytes, urea, and uric acid.

In patients with hyponatraemia (serum sodium < 134 mmol/l,reference range: 136–145 mmol/l), the detection of theunderlying pathogenetic mechanisms was our primary task.Diagnostic approach was based on the patients’ history andtheir physical examination, with emphasis on extracellularvolume status and on the correct interpretation of laboratorytests. The first step in this approach was to confirm the presenceof hypo-osmolality (Posm < 280 mOsm/kg) (Fig. 1). In thesetting of normal or elevated effective serum osmolality[measured Posm – blood urea (mg/dl)/6], evaluation of one ofthe causes of pseudohyponatraemia was carried out (severehyperlipidaemia or hyperproteinaemia) (Weisberg, 1989;Rose, 1994). Once it was demonstrated that the patient washypo-osmolar, the evaluation of the Uosm was used to deter-mine whether water excretion was normal or impaired. A valueof Uosm < 100 mOsm/kg was indicative of a complete and

© 2000 Medical Council on Alcoholism

612

Alcohol & Alcoholism Vol. 35, No. 6, pp. 612–616, 2000

MECHANISMS OF HYPONATRAEMIA IN ALCOHOL PATIENTS

GEORGE L. LIAMIS, HARALAMPOS J. MILIONIS, EVANGELOS C. RIZOS, KOSTAS C. SIAMOPOULOS and MOSES S. ELISAF*

Department of Internal Medicine, Medical School, University of Ioannina, 45110 Ioannina, Greece

(Received 21 March 2000; in revised form 5 June 2000; accepted 13 June 2000)

Abstract — Hyponatraemia is commonly reported in chronic alcoholic patients. However, the underlying pathogenetic mechanismsare not well delineated. In the current study, we analysed the possible pathophysiological mechanisms of hyponatraemia in a group ofalcoholic patients (n = 127) admitted to our hospital for causes related to alcohol misuse. Hyponatraemia (serum sodium <134 mmol/l)was found in 22 patients (17.3%). The most common cause of hyponatraemia in our cohort was hypovolaemia (12 patients); pseudo-hyponatraemia was diagnosed in six patients with alcohol-induced severe hypertriglyceridaemia. It is of interest that two patientsfulfilled the criteria of the so-called ‘beer potomania’ syndrome, while in two others, hyponatraemia was due to reset osmostat or tocerebral salt wasting syndrome, not previously described in alcoholic patients. It is concluded that hyponatraemia is a frequentlyobserved electrolyte disorder in hospitalized alcoholic patients and is related to various pathophysiological mechanisms.

*Author to whom correspondence should be addressed.

Table 1. Causes of admission of patients studied (n = 127)

Reason No. %

Alcohol withdrawal syndrome 25 19.7Increased serum liver enzymes 27 21.2and/or hepatomegalyAcute intoxication 26 20.5Anaemia 11 8.7Diarrhoea 7 5.5Chronic alcoholic pancreatitis 7 5.5Gastrointestinal symptoms (nausea, vomiting, 6 4.7dyspepsia, epigastric pain)Chronic myopathy 4 3.1Peripheral neuropathy 6 4.7Epistaxis 3 2.4Ataxia 5 3.9

appropriate suppression of antidiuretic hormone (ADH)secretion, a finding seen with either primary polydipsia(including ‘beer potomania’) or reset osmostat (Gillum andLinas, 1984). The criteria used to diagnose the ‘beer potomania’syndrome included a history of binge beer drinking and poordietary intake along with decreased serum sodium levels in the absence of other known causes of hyponatraemia (Hildenand Svendsen, 1975; Fenves et al., 1996). When Uosm exceeded100 mOsm/kg, hyponatraemia due to impaired water excretionwas diagnosed. Since patients with ascites or peripheral oedemawere excluded, normovolaemic or hypovolaemic hyponatraemia

was established. In such cases, the urine sodium was evaluatedin addition to assessing adrenal and thyroid function. Takinginto account that patients with renal failure as well as patientsreceiving diuretics were excluded, a urine sodium less than 20 mmol/l was indicative of hypovolaemia, while a urine sodiumgreater than 40 mmol/l was suggestive of the syndrome ofinappropriate antidiuresis (SIADH), reset osmostat, and of saltwasting conditions (Chung et al., 1987). In patients withequivocal findings (urine sodium 20–40 mmol/l), the correctdiagnosis was based on the response of serum sodium levelsfollowing the i.v. administration of sodium chloride (2 l of NaCl

HYPONATRAEMIA IN ALCOHOLICS 613

Fig. 1. Flow diagram for the diagnosis of hyponatraemia.CH2O, free water clearance; SIADH, syndrome of inappropriate antidiuresis.

0.9% w/v/day for 2 days). Hypovolaemia was diagnosed if theserum sodium concentration increased by 5 mmol/l or more asa response to this test (Chung et al., 1987). The reset osmostatsyndrome was suspected in patients with persistenthyponatraemia accompanied by an appropriate decrease in Uosm(<100 mOsm/kg) but inappropriate natriuresis (>40 mmol/l).In this setting, the diagnosis was confirmed by calculatingfree-water clearance (CH2O) (Defronzo et al., 1976; Elisaf et al.,1996). Finally, the changes of serum uric acid levels and uricacid FE were evaluated in the differential diagnosis betweensalt wasting syndrome and SIADH. The restoration of the serumsodium levels by salt supplementation and fluid restrictionfollowed by a reduction in urinary urate excretion pointed tothe diagnosis of SIADH, whereas the persistence of hypo-uricaemia and uric acid wasting following normalization ofserum sodium levels was suggestive of salt wasting (Maesaka,1996).

RESULTS

The acid-base and electrolyte abnormalities of the studypopulation are shown in Table 2. Twenty-two of the 127 patientshad hyponatraemia with a range of serum sodium between 121 and 133 mmol/l. However, no symptoms related tohyponatraemia were evident. The causes of hyponatraemiaaccording to the diagnostic work-up are shown in Table 3.Thus, six patients had pseudohyponatraemia due to alcohol-induced severe hypertriglyceridaemia (serum triglycerides2000–7200 mg/dl). A decrease in serum triglycerides wasfollowed by the restoration of serum sodium levels in all cases. The most common cause of hyponatraemia in our cohortwas hypovolaemia. In these patients, fluid loss was confirmedby history (e.g. vomiting and diarrhoea) and laboratory findings

suggestive of hypovolaemia, such as an increased serumurea/creatinine ratio (>40), low urine sodium concentration(<40 mmol/l), low FENa+ (<0.5%) and low FE of urea (<55%)were detected. Normalization of serum sodium levels wasachieved following infusion of normal saline with a greater than5 mmol/l increase in serum sodium levels observed after a 2-day sodium chloride loading.

Two patients with serum sodium levels of 122 and 124 mmol/l,respectively, had the so-called ‘beer potomania’ syndrome.The patients were drinking huge quantities of beer and theirdiet contained very small quantities of proteins and salt. Inboth patients, urine sodium concentration was extremely low (<10 mmol/l), Uosm was appropriately low (90 and 100 mOsm/kg) and potassium levels were in the lower normallimits (3.5 and 3.6 mmol/l, respectively). In one malnourishedpatient with low serum albumin levels (29 g/l), the diagnosis of the reset osmostat syndrome was established. This patient had asymptomatic hyponatraemia (serum sodium 128 mmol/l)with inappropriate natriuresis (urine sodium 64 mmol/l) butappropriately low Uosm (95 mOsm/kg) and excreted 85% of astandard water load (20 ml/kg body weight) within 4 h. Onphysical examination, peripheral oedema was not present andno postural changes in blood pressure and pulse rate wereevident. Thyroid, renal, liver, and adrenal function werenormal. Sodium chloride administration was not followed by an increase in serum sodium levels, while sodium balancewas preserved. The maximum free water clearance (CH2O) was10.5 ml/min and did not differ from observations in previousstudies on healthy individuals (reference range 5.8–15 ml/min)(Cooke et al., 1979). Hyponatraemia due to cerebral saltwasting syndrome was diagnosed in a 68-year-old patient with dementia in association with alcohol-induced cerebralatrophy. Initially, the diagnosis of SIADH was made based onthe coexistence of hyponatraemia (serum sodium 122 mmol/l)with inappropriate natriuresis (urine sodium 56 mmol/l) andhypouricaemia with uricosuria [serum uric acid levels 3.2 mg/dl(reference range 4–7.2 mg/dl) and FE of uric acid 18% (refer-ence range 11–16%)]. However, the diagnosis of cerebral salt-wasting syndrome was established after a careful diagnosticfollow-up. In fact, despite the natriuresis, there was evidencefor volume depletion (postural hypotension). Furthermore, the correction of hyponatraemia by salt supplementation andfluid restriction was not associated with a significant rise inserum uric acid levels (serum uric acid levels 3.4 mg/dl) or areduction in FE of uric acid (FE of uric acid 17.5%).

DISCUSSION

In the present study, the prevalence of hyponatraemia andits associated pathogenetic mechanisms were assessed in agroup of hospitalized chronic alcoholic patients. In our series,hyponatraemia was the third commonest electrolyte abnormalitydetected (17.3%). However, in a considerable number of thesepatients, pseudohyponatraemia was diagnosed due to alcohol-induced hypertriglyceridaemia. In such cases, there is amarked increase in the non-aqueous phase of plasma resultingin a fall in the ratio of sodium to total volume, even though thesodium in the aqueous phase is normal and so is Posm (Albrinket al., 1955). True volume depletion mainly due to gastro-intestinal fluid losses represents the chief cause of hyponatraemia

614 G. L. LIAMIS et al.

Table 2. Acid-base and electrolyte abnormalities in the population studied (n = 127)

Abnormality No. %

Respiratory alkalosis (arterial pH > 7.46) 20 15.7Metabolic alkalosis (arterial pH > 7.46) 7 5.5Acidaemiaa (arterial pH < 7.34) 14 11Hyponatraemia (serum sodium < 134 mmol/l) 22 17.3Hypernatraemia (serum sodium > 146 mmol/l) 3 2.3Hypokalaemia (serum potassium < 3.4 mmol/l) 16 12.6Hypomagnesaemia (serum magnesium 38 29.9< 0.65 mmol/l)Hypophosphataemia (serum phosphorus 37 29.1< 0.77 mmol/l)Hypocalcaemia (serum calcium < 2.1 mmol/l) 26 (18)b 20.5 (14.2)b

Hyperphosphataemia (serum phosphorus 2 1.6> 1.45 mmol/l)

aDue to either pure metabolic acidosis or to mixed acid–base disorders;b18 patients with true hypocalcaemia (14.2%).

Table 3. Observed causes of hyponatraemia

Cause No. %

Pseudohyponatraemia 6 27.3Hypovolaemia 12 54.5‘Beer’ potomania syndrome 2 9.1Cerebral salt-wasting syndrome 1 4.5Reset osmostat syndrome 1 4.5

in alcoholic patients. Effective volume depletion predisposestoward the development of hyponatraemia by its effects onrenal water excretion and thirst. In fact, hypovolaemia is apotent stimulus to ADH secretion, resulting in water retention.Furthermore, volume depletion can directly or indirectlystimulate thirst and consequently increase water intake (Rose,1994). The diagnosis of hypovolaemic hyponatraemia issometimes evident from history and physical examination.However, the accuracy of clinical evaluation for predicting the state of extracellular fluid volume in hyponatraemia is low. Laboratory markers, such as the increased serum urea/creatinine ratio and the low urine sodium concentration, areparticularly useful in this setting. In doubtful cases, the responseof the serum sodium concentration to the i.v. administration ofnormal saline for 2 days is useful in establishing the correctdiagnosis. An increase in serum sodium concentration greaterthan 5 mmol/l is indicative of hypovolaemic hyponatraemia,which was the case in all hypovolaemic patients (Chung et al.,1987). It has been proposed that the combination of low FE ofsodium (<0.5%) and low FE of urea (<55%) can safely predictsaline response (Musch et al., 1995). In fact, in our series, allpatients with hypovolaemic hyponatraemia satisfied thesediagnostic criteria. The presence of low sodium excretion inalcoholic patients strongly favours the diagnosis of hypo-volaemia, since these patients commonly exhibit reversibletubular injury as a result of alcoholism and tend to increasesolute excretion because of the potential diuretic effect ofpoorly reabsorbed substances, such as ketoacids or lactate (De Marchi et al., 1993).

The syndrome of ‘beer potomania’ was diagnosed in twopatients. These patients were malnourished binge drinkers andpresented with profound hyponatraemia. In ‘beer potomania’,hyponatraemia is due to a large consumption of beer (whichhas a poor salt content) together with a minimal intake ofordinary food. A diet poor in salt and protein compounds (i.e. urea precursors) results in reduced excretion of urinarysolutes, which limits the ability to excrete free water (Hildenand Svendsen, 1975; Fenves et al., 1996; Blaustein andSchwenk, 1997). In both patients, there was an appropriaterenal response to the large fluid intake by maximally dilutingthe urine (Uosm ≤ 100 mOsm/kg). Binge drinkers typicallyproduce less than 250 mOsm of solutes a day; hence, ourpatients could probably excrete no more than 2.5 l of urinedaily (250/100 = 2.5 l). Since they were drinking about 3 l of beer daily, about 0.5 l of fluid was retained leading todilutional hyponatraemia.

In one patient presenting hyponatraemia with inappropriatenatriuresis, the syndrome of SIADH was implicated. However,the patient fulfilled the criteria for the diagnosis of the resetosmostat syndrome (Defronzo et al., 1976; Wall et al., 1992;Elisaf et al., 1996). Specifically, this patient had:

(1) normovolaemic hyponatraemia and could maintainsodium balance without correcting the existing hyponatraemiawhen sodium chloride intake was increased; (2) an intacturinary diluting ability, as his urine osmolality was lower than100 mOsm/kg; (3) normal excretion of a standard water load(>80% within 4 h); (4) normal adrenal, renal, and thyroidfunction together with no evidence of cardiac or hepatic dis-ease. Nevertheless, we cannot disregard the fact that this patientmay have had a mild degree of volume depletion. Thus, ADHsecretion adjusted for hypo-osmolality as well as (to some

extent) for hypovolaemia could be responsible for the patient’sclinical presentation.

Hyponatraemia due to reset osmostat has not been reportedin alcoholic patients. There is evidence that this syndrome ispresent in approximately one-third of patients with SIADHand may occur in several other conditions, including chronicmalnutrition, as was the case in our patient (Defronzo et al.,1976). In this setting, defective cellular metabolism may beresponsible for the abnormal osmoreceptor function. Inter-estingly, correction of the underlying problem and hyper-alimentation were effective in returning the serum sodiumconcentration towards normal.

Finally, one patient with alcohol-induced cerebral atrophyand dementia presented with hyponatraemia related tocerebral salt-wasting syndrome. The diagnosis was based on arecently proposed diagnostic work-up taking into account uricacid metabolism and the patient’s response to saline infusion(Maesaka, 1996). It is of interest that patients with this syn-drome develop hyponatraemia with all the SIADH-associatedfindings (including inappropriate natriuresis and hypo-uricaemia). However, these patients are volume depleted andhave a high urine sodium concentration due to urinary sodiumwasting and not to volume expansion (Al Mufti and Arief,1984; Ishikawa et al., 1987; Tanneau et al., 1987). The correc-tion of hyponatraemia after saline infusion and the persistenceof hypouricaemia (and inappropriate uric acid wasting) despitenormalization of sodium levels suggest salt wasting (Maesaka,1996). The cerebral salt-wasting syndrome is observed inpatients with cerebral diseases and its pathogenesis remainsunclear. It has been proposed that there is an increased releaseof natriuretic peptides from specific hormone-producingneurons in the brain stimulated by disorders of the centralnervous system (Tanneau et al., 1987).

In conclusion, hyponatraemia is a frequently observedelectrolyte abnormality in hospitalized chronic alcoholicpatients and is related to various pathophysiologicalmechanisms.

REFERENCES

Albrink, M. J., Hald, P. M., Man, E. B. and Peters, J. P. (1995) Thedisplacement of serum water by the lipids of hyperlipemic serum. A new method for the rapid determination of serum water. Journalof Clinical Investigation 34, 1483–1488.

Al Mufti, H. and Arieff, A. I. (1984) Hyponatremia due to cerebral salt-wasting syndrome. Combined cerebral and distal tubular lesion.American Journal of Medicine 77, 740.

Blaustein, D. A. and Schwenk, M. H. (1997) “Beer potomania” in anon-beer drinker Clinical Nephrology 48, 203–204.

Chung, H. M., Kluge, R., Schrier, R. W. and Anderson, R. J. (1987)Clinical assessment of extracellular volume in hyponatremia.American Journal of Medicine 83, 905–908.

Cooke, C. R. Turin, M. D. and Walker, W. G. (1979) The syndrome ofinappropriate antidiuretic hormone secretion (SIADH): patho-physiologic mechanisms in solute and volume regulation. Medicine58, 240–251.

Defronzo, R. A., Goldberg, M. and Agus, Z. S. (1976) Normal dilutingcapacity in hyponatremic patients: Reset osmostat or a variant of thesyndrome of inappropriate antidiuretic hormone secretion. Annals ofInternal Medicine 84, 538–542.

De Marchi, S., Cecchin, E., Basile, A., Bertotti, A., Nardini, R. andBartoli, E. (1993) Renal tubular dysfunction in chronic alcohol abuse— Effects of abstinence. New England Journal of Medicine 329,1927–1934.

HYPONATRAEMIA IN ALCOHOLICS 615

616 G. L. LIAMIS et al.

Elisaf, M., Merkouropoulos, M., Tsianos, E. V. and Siamopoulos, K. C.(1994) Acid-base and electrolyte abnormalities in alcoholic patients.Mineral and Electrolyte Metabolism 20, 274–281.

Elisaf, M., Konstantinides, A. and Siamopoulos, K. C. (1996) Chronichyponatremia due to reset osmostat in a patient with colon cancer.American Journal of Nephrology 16, 349–351.

Fenves, A. Z., Thomas, S. and Knochel, J. P. (1996) Beer potomania:two cases and review of the literature. Clinical Nephrology 45, 61–64.

Gillum, D. M. and Linas, S. L. (1984) Water intoxication in a psychoticpatient with normal water excretion. American Journal of Medicine77, 773–774.

Hilden, T. and Svendsen, T. L. (1975) Electrolyte disturbances in beerdrinkers. A specific hypoosmolality syndrome. Lancet ii, 245–246.

Ishikawa, S-E., Saito, T., Kaneko, T., Okada, K. and Kuzuya, T. (1987)Hyponatremia responsive to flurocorticosterone acetate in elderlypatients after head injury. Annals of Internal Medicine 106, 187–191.

Kaysen, G. and North, R. H. (1984) The effects of alcohol on blood pres-sure and electrolytes. Medical Clinics of North America 68, 221–246.

Maesaka, J. K. (1996) An expanded view of SIADH, hyponatremia andHypouricemia. Clinical Nephrology 46, 79–83.

Musch, V., Thimpont, J., Vandervelde, D. Verhaeverbeke, I.,Berghmans, T. and Decaux, G. (1995) Combined fractionalexcretion of sodium and urea better predicts response to saline inhyponatremia than do usual clinical and biochemical parameters.American Journal of Medicine 99, 348–355.

Rose, B. D. (1994) Hypoosmolar states — Hyponatremia. In ClinicalPhysiology of Acid-base and Electrolyte Disorders, 4th edn, pp. 651–694. McGraw-Hill, New York.

Schaefer, R. M., Teschner, M. and Heidland, A. (1987) Alterations of water, electrolyte and acid-base homeostasis in the alcoholic.Mineral and Electrolyte Metabolism 13, 1–6.

Tanneau, R. A., Pennec, Y. L., Jouquan, J. and Le Menn, G. (1987)Cerebral salt-wasting in elderly patients. Annals of InternalMedicine 107, 120.

Wall, R. M., Crofton, T. J. M., Share, L. and Cooke, C. R. (1992)Chronic hyponatremia due to resetting of the osmostat in a patientwith gastric carcinoma. American Journal of Medicine 93,223–228.

Weisberg, L. S. (1989) Pseudohyponatremia. A reappraisal. AmericanJournal of Medicine 86, 315–318.